No abstract

Integrating humans and robotic machines into one system offers multiple opportunities for creating assistive technologies that can be used in biomedical, industrial, and aerospace applications. The scope of the present research is to study the integration of a human arm with a powered exoskeleton (orthotic device) and its experimental implementation in an elbow joint, naturally controlled by the human. The Human-Machine interface was set at the neuromuscular level, by using the neuromuscular signal (EMG) as the primary command signal for the exoskeleton system. The EMG signal along with the joint kinematics were fed into a myoprocessor (Hill-based muscle model) which in turn predicted the muscle moments on the elbow joint. The moment-based control system integrated myoprocessor moment prediction with feedback moments measured at the human arm/exoskeleton and external load/exoskeleton interfaces. The exoskeleton structure under study was a two-link, two-joint mechanism, corresponding to the arm limbs and joints, which was mechanically linked (worn) by the human operator. In the present setup the shoulder joint was kept fixed at given positions and the actuator was mounted on the exoskeleton elbow joint. The operator manipulated an external weight, located at the exoskeleton tip, while feeling a scaled-down version of the load. The remaining external load on the joint was carried by the exoskeleton actuator. Four indices of performance were used to define the quality of the human/machine integration and to evaluate the operational envelope of the system. Experimental tests have shown that synthesizing the processed EMG signals as command signals with the external-load/human-arm moment feedback, significantly improved the mechanical gain of the system, while maintaining natural human control of the system, relative to other control algorithms that used only position or contact forces. The results indicated the feasibility of an EMG-based power exoskeleton system as an integrated human-machine system using high-level neurological signals.

The Raven-II is a platform for collaborative research on advances in surgical robotics. Seven universities have begun research using this platform. The Raven-II system has two 3-DOF spherical positioning mechanisms capable of attaching interchangeable four DOF instruments. The Raven-II software is based on open standards such as Linux and ROS to maximally facilitate software development. The mechanism is robust enough for repeated experiments and animal surgery experiments, but is not engineered to sufficient safety standards for human use. Mechanisms in place for interaction among the user community and dissemination of results include an electronic forum, an online software SVN repository, and meetings and workshops at major robotics conferences.

The best method of training for laparoscopic surgical skills is controversial. Some advocate observation in the operating room, while others promote animal and simulated models or a combination of surgery-related tasks. A crucial process in surgical education is to evaluate the level of surgical skills. For laparoscopic surgery, skill evaluation is traditionally performed subjectively by experts grading a video of a procedure performed by a student. By its nature, this process uses fuzzy criteria. The objective of the current study was to develop and assess a skill scale using Markov models (MMs). Ten surgeons [five novice surgeons (NS); five expert surgeons (ES)] performed a cholecystectomy and Nissen fundoplication in a porcine model. An instrumented laparoscopic grasper equipped with a three-axis force/torque (F/T) sensor was used to measure the forces/torques at the hand/tool interface synchronized with a video of the tool operative maneuvers. A synthesis of frame-by-frame video analysis and a vector quantization algorithm, allowed to define F/T signatures associated with 14 different types of tool/tissue interactions. The magnitude of F/T applied by NS and ES were significantly different (p < 0.05) and varied based on the task being performed. High F/T magnitudes were applied by NS compared with ES while performing tissue manipulation and vise versa in tasks involved tissue dissection. From each step of the surgical procedures, two MMs were developed representing the performance of three surgeons out of the five in the ES and NS groups. The data obtained by the remaining two surgeons in each group were used for evaluating the performance scale. The final result was a surgical performance index which represented a ratio of statistical similarity between the examined surgeon's MM and the MM of NS and ES. The difference between the performance index value, for a surgeon under study, and the NS/ES boundary, indicated the level of expertise in the surgeon's own group. Using this index, 87.5% of the surgical procedures were correctly classified into the NS and ES groups. The 12.5% of the procedures that were misclassified were performed by the ES and classified as NS. However in these cases the performance index values were very close to the NS/ES boundary. Preliminary data suggest that a performance index based on MM and F/T signatures provides an objective means of distinguishing NS from ES. In addition, this methodology can be further applied to evaluate haptic virtual reality surgical simulators for improving realism in surgical education.

Accurate knowledge of biomechanical characteristics of tissues is essential for developing realistic computer-based surgical simulators incorporating haptic feedback, as well as for the design of surgical robots and tools. As simulation technologies continue to be capable of modeling more complex behavior, an in vivo tissue property database is needed. Most past and current biomechanical research is focused on soft and hard anatomical structures that are subject to physiological loading, testing the organs in situ. Internal organs are different in that respect since they are not subject to extensive loads as part of their regular physiological function. However, during surgery, a different set of loading conditions are imposed on these organs as a result of the interaction with the surgical tools. Following previous research studying the kinematics and dynamics of tool/tissue interaction in real surgical procedures, the focus of the current study was to obtain the structural biomechanical properties (engineering stress-strain and stress relaxation) of seven abdominal organs, including bladder, gallbladder, large and small intestines, liver, spleen, and stomach, using a porcine animal model. The organs were tested in vivo, in situ, and ex corpus (the latter two conditions being postmortem) under cyclical and step strain compressions using a motorized endoscopic grasper and a universal-testing machine. The tissues were tested with the same loading conditions commonly applied by surgeons during minimally invasive surgical procedures. Phenomenological models were developed for the various organs, testing conditions, and experimental devices. A property database-unique to the literature-has been created that contains the average elastic and relaxation model parameters measured for these tissues in vivo and postmortem. The results quantitatively indicate the significant differences between tissue properties measured in vivo and postmortem. A quantitative understanding of how the unconditioned tissue properties and model parameters are influenced by time postmortem and loading condition has been obtained. The results provide the material property foundations for developing science-based haptic surgical simulators, as well as surgical tools for manual and robotic systems.